CN-122018181-A - Underground protective glasses capable of achieving eye movement and multi-mode sensing and early warning method

CN122018181ACN 122018181 ACN122018181 ACN 122018181ACN-122018181-A

Abstract

The invention relates to the technical field of underground safety monitoring and occupational health protection, and discloses underground protective glasses and an early warning method for eye movement and multi-mode perception, wherein the glasses integrate a main control unit in a glasses frame and glasses leg assembly, the device is connected with an elevation angle eye sensing module, an inertial sensor, a PPG and body temperature piezoelectric sensor, an acoustic assembly, a wireless communication module and a detachable glasses leg battery module. The method is characterized by constructing a multidimensional state discrimination model by collecting eye images and multisource physiological data in real time, creatively utilizing the logic contradiction of normal blood oxygen values and slow eye movement reaction to identify hidden carbon monoxide poisoning, distinguishing accidental slip and coma falling through a gesture and eye movement cooperative mechanism, and simultaneously having an active noise reduction function. The invention solves the problems that the traditional equipment cannot identify hidden poisoning and has high false alarm rate of falling, and realizes accurate perception and early warning of physiological fatigue, toxic symptoms and safety states of underground personnel.

Inventors

LIU YEJIAO
JIANG FENGYI
LIU SHUHAN
TIAN ZHICHAO
LI MIAO
WANG HUIXIN
GAO FU
LI JINLIANG
YAN WENJIE

Assignees

内蒙古科技大学

Dates

Publication Date: 20260512
Application Date: 20260127

Claims (10)

1. The downhole protective glasses with eye movement and multi-mode perception comprise a glasses frame and a glasses leg assembly (2), and are characterized in that an embedded main control unit (3) is arranged in the glasses frame and the glasses leg assembly (2), an eye movement sensing module (1), an inertial sensor, a reflective PPG sensor (4), a body temperature and skin electric sensor (5) and a pickup microphone (6) are respectively connected to the input end of the embedded main control unit (3), and a loudspeaker (7) and a wireless communication module are connected to the output end of the embedded main control unit (3); The eye sensing module (1) is arranged at the nose pad area or the lower edge of the glasses frame and the glasses leg assembly (2) and is configured to collect image data of eyes by using an elevation angle light path; the inertial sensor is integrated in the mirror frame and the mirror leg assembly (2) and is configured to collect head movement posture data; The reflective PPG sensor (4) and the body temperature and skin electric sensor (5) are arranged on the inner side wall of the glasses frame and the glasses leg assembly (2), and the positions of the reflective PPG sensor and the body temperature and skin electric sensor correspond to the temple or the behind-the-ear area of a wearer and are configured to collect physiological signals; The pickup microphone (6) and the loudspeaker (7) are respectively arranged at the outer sides of the glasses frame and the glasses leg assembly (2) and at positions corresponding to the auditory meatus; The embedded main control unit (3) is configured to perform logic determination of physiological fatigue, environmental poisoning and falling accidents based on the acquired data and control the horn (7) to alarm.
2. The downhole glasses according to claim 1, wherein the eye-sensing module (1) comprises a micro-infrared CMOS image sensor and a near-infrared light-compensating lamp, the near-infrared light-compensating lamp is configured to emit near-infrared light having a wavelength range of 750 nm to 1000 nm, the micro-infrared CMOS image sensor is configured to operate in a global exposure mode, and an optical axis of the micro-infrared CMOS image sensor is inclined upward with respect to a horizontal line of sight, avoiding a forward-looking line area of the wearer.
3. The downhole glasses according to claim 1, wherein the embedded main control unit (3) is further communicatively connected with an environment sensing interface configured to obtain carbon monoxide concentration and gas concentration data in the environment, and the environment sensing interface is connected with a micro gas sensor integrated in the glasses frame and the glasses leg assembly (2) or with a portable gas detector carried by a wearer.
4. The downhole glasses according to claim 1, wherein the end of the frame and temple assembly (2) is provided with a detachable temple battery module (8); the detachable glasses leg battery module (8) is connected through a magnetic contact or a buckling structure, and is arranged in a position located in a region behind the ear of a wearer, and forms a gravity balance weight for the front eye sensing module (1); and a standby power supply is integrated on a circuit board where the embedded main control unit (3) is located and is configured to maintain system operation to support hot plug when the detachable mirror leg battery module (8) is disconnected.
5. An eye movement and multi-mode sensing downhole pre-warning method, which is applied to the eye movement and multi-mode sensing downhole protective glasses according to any one of claims 1 to 4, and comprises the following steps: S1, acquiring and preprocessing multi-source heterogeneous data in real time, wherein an embedded main control unit (3) synchronously acquires eye image data of an eye sensing module (1), photoelectric volume pulse wave signals of a reflective PPG sensor (4), skin electric response signals of a body temperature and a skin electric sensor (5), head movement posture data of an inertial sensor and environmental gas concentration data, and carries out filtering and denoising preprocessing on the acquired data; S2, deeply extracting eye features and physiological features, namely carrying out feature calculation on the preprocessed data by the embedded main control unit (3), and extracting a PERCLOS value, a pupil diameter, a pupil response speed, an eyeball saccade speed, a blood oxygen saturation value, a heart rate variability index and a skin conductivity change trend; s3, constructing a three-dimensional state discrimination model of eye movement characteristics, blood oxygen data and environmental data and carrying out logic judgment, wherein the embedded main control unit (3) inputs the characteristic parameters extracted in the step S2 into a preset logic discrimination model to judge whether an operator is in a pure physiological fatigue state, an anoxic or dominant poisoning state or a hidden carbon monoxide poisoning state; S4, fall detection and false alarm rejection based on coordination of gestures and eye movements, wherein the embedded main control unit (3) calculates a combined acceleration value based on the head movement gesture data and identifies fall time sequence characteristics, and after judging that a fall event occurs, the synchronous eye movement characteristic data is used for judging the conscious state of a wearer so as to distinguish accidental slipping from comatose falling; And S5, executing active hearing protection, grading early warning and data reporting, wherein the embedded main control unit (3) controls the loudspeaker (7) to send out alarm prompts of corresponding grades according to the judging results of the step S3 and the step S4, reports a distress data packet through the wireless communication module, and utilizes data acquired by the pickup microphone (6) to generate inverted sound waves to perform active noise reduction.
6. The downhole pre-warning method of eye movement and multi-modal sensing according to claim 5, wherein in step S3, the decision logic of the hidden carbon monoxide poisoning state is: When the blood oxygen saturation value is larger than or equal to a preset normal blood oxygen lower limit threshold value, the saccade speed of the eyeball is smaller than a preset nerve reaction retardation threshold value, and the carbon monoxide concentration value in the environmental gas concentration data is larger than the minimum detection limit of the sensor, the embedded main control unit (3) recognizes the logic contradiction between the physiological blood oxygen reading display normal and nerve reaction state display inhibition, and judges the state to be a concealed carbon monoxide poisoning state.
7. The downhole pre-warning method according to claim 5, wherein in step S4, the misinformation rejection and consciousness state confirmation mechanism comprises: when detecting the time sequence change of the total acceleration conforming to the falling characteristic, the embedded main control unit (3) reads the eye movement characteristic data of the falling resting stage; if the eye movement characteristic data show that the saccade speed of the eyeball is greater than a preset active threshold value or the blink frequency is in a normal range, judging that the eyeball is consciously suddenly slips, and not triggering emergency help seeking; If the eye movement characteristic data show that the eye saccade speed is approaching zero, the eyelid is closed for a long time or the pupil is disappeared from light reflection, the eye movement characteristic data are judged to be coma or severe injury and fall, and emergency help seeking is triggered.
8. The downhole pre-warning method of eye movement and multi-modal sensing according to claim 5, wherein in step S1, the preprocessing of the photoplethysmography signal comprises: The embedded main control unit (3) uses the synchronously acquired head motion gesture data as a reference signal, and adopts a least mean square adaptive filtering algorithm to estimate and subtract a noise component related to head motion from an original photoplethysmography pulse wave signal.
9. The downhole pre-warning method according to claim 5, wherein in step S5, the active hearing protection is performed by: The embedded main control unit (3) performs spectrum analysis on sound signals collected by the pickup microphone (6); identifying a frequency spectrum component with a frequency lower than a preset low frequency threshold value and a stable amplitude as steady-state low-frequency noise; calculating an inverted sound wave signal with the same frequency, the same amplitude and opposite phase to the steady-state low-frequency noise; And the inverse sound wave signal and the voice and alarm signal subjected to transparent transmission processing are digitally synthesized and then played through the loudspeaker (7).
10. The downhole pre-warning method of eye movement and multi-modal sensing according to claim 5, wherein in step S2, the calculation method of the PERCLOS value is: Positioning the upper and lower edge coordinates of the eyelid by using a face key point detection algorithm and calculating the eyelid opening amplitude; counting the accumulated duration of the eye closure frames with eyelid opening amplitude less than twenty percent of full opening amplitude in a preset time window; Dividing the accumulated duration by the total duration of the preset time window to obtain the PERCLOS value.

Description

Underground protective glasses capable of achieving eye movement and multi-mode sensing and early warning method Technical Field The invention relates to the technical field of underground safety monitoring and occupational health protection, in particular to underground protective glasses and an early warning method for eye movement and multi-mode perception. Background At present, the underground coal mining environment is extremely complex and severe, and the operation safety is always the core focus of industry attention. Through the deep disc-copying discovery of the reasons of past accidents, unsafe behaviors of people are main causes of the accidents. The physiological and psychological states of the operators, especially the fatigue degree, the concentration degree and the consciousness definition, directly determine the normalization and the accuracy of the operation of the equipment. Therefore, real-time individual state monitoring of underground operators has become a key link for preventing human misoperation and reducing accident risk from the source. Aiming at the requirement on personnel state monitoring, the existing application scheme mostly adopts a non-contact machine vision technology. The common implementation mode is to integrate the camera module at the cap peak of the mine safety helmet or fix the camera module on the mine lamp module in a hanging way. When the system works, video stream data of the face of a wearer are continuously collected, an eye region is positioned by using an image processing algorithm, and the duty ratio (PERCLOS value) of the eyelid closing duration in a certain time is calculated. When the value exceeds a preset threshold, the system judges that the personnel is in a fatigue sleepy state and triggers a local audible and visual alarm or uploads alarm information through a base station. However, such prior art has limitations in practical applications. Firstly, the monitoring equipment is hung on a safety helmet or a miner lamp in an externally-hung mode, the load on the head is increased, the gravity center is changed, the wearing discomfort is caused, so that miners generate strong interference emotion, in addition, the head frequently shakes during operation, facial features are unstable to capture, and false alarms frequently occur. Secondly, a discrimination blind area exists in single dependence on eyelid closure index. Under the underground specific environment, the personnel reaction is slow, not only because of simple sleep deficiency, but also central nervous inhibition caused by low-concentration carbon monoxide poisoning or hypoxia; because carbon monoxide poisoning can make skin color appear cherry red, conventional pure visual monitoring or ordinary blood oxygen monitoring is difficult to detect the pathological abnormality, physiological fatigue and concealed poisoning cannot be distinguished, and comprehensive and effective life safety protection is difficult to provide. Therefore, the invention provides the underground protective glasses for eye movement and multi-mode sensing and the early warning method thereof, so as to solve the defects in the prior art. Disclosure of Invention Aiming at the defects of the prior art, the invention provides the underground protective glasses for eye movement and multi-mode perception and the early warning method, and solves the problems that the existing underground monitoring equipment is poor in wearing comfort, can not distinguish simple fatigue and hidden gas poisoning risks and lacks comprehensive protection functions of fall detection and hearing protection. In order to achieve the above purpose, the invention is realized by the following technical scheme: In a first aspect, the invention provides downhole protective glasses for eye movement and multi-mode perception, which comprise a glasses frame and a glasses leg assembly, wherein embedded main control units are arranged in the glasses frame and the glasses leg assembly; The input end of the embedded main control unit is respectively connected with an eye sensing module, an inertial sensor, a reflective PPG sensor, a body temperature and skin electric sensor and a pickup microphone, and the output end of the embedded main control unit is connected with a loudspeaker and a wireless communication module; the eye sensing module is arranged at the nose support area or the lower edge of the glasses frame and the glasses leg assembly and is configured to collect image data of eyes by using an elevation angle light path; the inertial sensor is integrated in the mirror frame and the mirror leg assembly and is configured to collect head movement posture data; The reflective PPG sensor and the body temperature and skin electric sensor are arranged on the inner side wall of the glasses frame and the glasses leg component, and a location corresponding to a temple or a behind-the-ear region of the wearer configured to acquire physiological signals; the pickup